PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
February 21, 2026Climate of the past0 citationsOpen Access

How temperature seasonality drives interglacial permafrost dynamics: implications for paleo reconstructions and future thaw trajectories

JNJan NitzbonAlfred-Wegener-Institut Helmholtz-Zentrum für Polar- und MeeresforschungMLMoritz LangerAlfred-Wegener-Institut Helmholtz-Zentrum für Polar- und MeeresforschungLMLuca Alexander Müller-IßbernerUniversity of Göttingen

Key Points

  • This study investigates the role of temperature seasonality in permafrost dynamics during interglacial periods, specifically focusing on the mid Holocene and last interglacial.
  • Combined paleo climate simulations using the AWI-ESM-2.5 model with the CryoGridLite permafrost model.
  • Analyzed ground thermal regimes and freeze-thaw dynamics in northern high-latitude areas.
  • Evaluated the influence of annual mean and seasonal temperature amplitudes on permafrost characteristics.
  • Simulations indicated lower temperatures during the mid Holocene and higher temperatures during the last interglacial compared to pre-industrial times.
  • Local and global permafrost extent was found to be primarily affected by mean temperatures.
  • Increased thaw depths correlate with both annual mean surface temperatures and seasonal amplitudes.

Abstract

Abstract. Various proxy records have suggested widespread permafrost degradation in northern high latitudes during interglacial warm climates, including the mid Holocene (MH, 6000 years before present) and the last interglacial (LIG, 127 ka BP), and linked this to substantially warmer high-latitude climates compared to the pre-industrial period (PI). However, most Earth system models suggest only modest warming or even slight cooling in terms of annual mean surface temperatures during these interglacials, seemingly contradicting the reconstructions of widespread permafrost degradation. Here, we combine paleo climate simulations of the Alfred Wegener Institute's Earth system model version 2.5 (AWI-ESM-2.5) with the CryoGridLite permafrost model to investigate the ground thermal regime and freeze-thaw dynamics in northern high-latitude land areas during the MH and the LIG in comparison to the PI. Specifically, we decompose how the annual mean and seasonal amplitude (that is, the difference between the maximum and minimum monthly mean) of surface temperatures affect the occurrence of permafrost, seasonal frost, thaw depth and duration, and thermal contraction cracking activity. For the MH (LIG) AWI-ESM-2.5 simulated global-mean surface temperatures in the simulation domain to be about 0.1 K lower (0.4 K higher), and the global-mean seasonal amplitudes to be 2.9 K (7.4 K) higher than for the PI. With respect to interglacial permafrost characteristics, our simulations revealed that (i) local permafrost probabilities and global permafrost extent are predominantly determined by mean temperatures, (ii) maximum thaw depths are increasing with both annual mean and seasonal amplitudes, and (iii) thermal contraction cracking within the permafrost domain is almost solely driven by the seasonal amplitudes of surface temperatures. Thus, not only mean warming, but also the enhanced seasonal temperature amplitude due to a different orbital forcing have driven permafrost and ground ice dynamics during past interglacial climates. Our results provide an additional explanation of reconstructed periods of marked permafrost degradation in the past, which was driven by deep surficial thaw during summer, while colder winters allowed for permafrost persistence in greater depths. Our results further suggest that past interglacial climates have limited suitability as analogues for future permafrost thaw trajectories, as rising mean temperatures paralleled by decreasing seasonal amplitudes expose the northern permafrost region to magnitudes of thaw that are likely unprecedented since at least Marine Isotope Stage 11c (about 400 ka BP).

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Nitzbon et al. (2026) studied this question.

synapsesocial.com/papers/69994c5d873532290d020bbehttps://doi.org/10.5194/cp-22-377-2026
Ask AI
Helpful
Bookmark
Share
View Full Paper